The subject of AI in healthcare covers a lot of ground. It can be hard to stay on top of all the latest advances in this evolutionary, if not revolutionary, technology space. Last week we examined some of the ways that technology tools are changing the game for health engagement. This week, we’ll look at how artificial intelligence in healthcare may affect our future as individuals and as part of a specific health population.

Some of the biggest names in healthcare are working hard to accelerate the use of AI in healthcare, whether it’s teaching robots how to mimic empathy, or a new AI bot that helps dermatologists screen patients for skin cancer. Read on to learn more about how big corporations like IBM, and a host of new technology startups, are literally changing the “face” of healthcare as we know it today.

#1 Trend –AI and Health Screenings

More and more, data gathered from health engagement programs proves that a stronger focus on wellness, including motivating individuals to get preventative screenings for diseases, can improve population health outcomes and lower healthcare costs overall.

More people are diagnosed every year with skin cancer than all other cancers combined, according to the American Cancer Society. Cases of melanoma, the deadliest form of skin cancer, have more than doubled in the past 30 years. Because of the rise in occurrences, dermatologists are constantly trying to find ways to detect skin cancer in its earliest form. When people live far from the doctor, regular skin screenings can become even more challenging for individuals and for health plans.

Enter a group of computer scientists at Stanford University using AI in healthcare to teach an algorithm how to diagnose and detect skin cancers. The researchers began with analgorithm developed by Google that was already trained to identify 1.28 million images from 1,000 object categories.

First, the researchers added tens of thousands of images of skin cancers and benign lesions. Then, they teamed up with Stanford dermatologists to develop a “sensitivity curve” that measures the accuracy of detection by the algorithm for both benign and malignant skin cancers. After a substantial “learning” period, the algorithm can now perform as well as the actual dermatologists in detecting skin cancer.

The Stanford team is not trying to replace the work of actual dermatologists. By creating algorithms like these, however, telehealth applications using smartphones and other remote devices may be as effective as seeing a doctor in person. And that could be one AI in healthcare trend that might change health outcomes in a dramatically positive direction.

#2 Trend – Deep Learning and Early Diagnosis

AI in healthcare is already working hard to do more than diagnose skin cancer alone. A form of machine learning called “deep learning” is also helping researchers find other diseases much earlier than in the past. Basically, deep learninginvolves feeding systems more and more data through neural networks. The computer system can then use that data to make decisions about other data.

To understand this from a consumer perspective, when Amazon and Netflix help you decide what you’d like to buy or watch next, the algorithms are using a form of deep learning to help determine that information.

Within healthcare, deep learning can be used to help computer systems become increasingly intelligent about the signs and symptoms of life-threatening diseases. In many cases this even involves our sense of smell. For example, an Israeli chemical engineer has invented a “smelling machine” thatuses odors to distinguish 17 different diseases with up to 86 percent accuracy. As the system is exposed to more and more smells that are associated with cancers and other illnesses, it becomes more accurate as it “learns.”

Millions of dollars have been raised to pursue “odor analytics technology” involving the development of sensors that can detect the smells of significant illnesses.Although this might put some cancer-sniffing dogs out of business, the application of AI in healthcare and deep learning to the task of early diagnosis of cancers and other life-threatening diseases could take preventative health screenings to a whole new level.

Rise of Health Action Platform

Executive Brief

See for yourself how a flexible health action technology platform can lower costs and evolve to meet the changing needs of healthcare organizations and consumers.

Get Report

#3 Trend – Data and Doctors

AI in healthcare is not only advancing earlier diagnoses, there are doctors using this technology to analyze data to help improve treatment programs.

Doctors at Memorial Sloane Kettering have been working with data analysts for over five years to help teach IBM’s Watson for Oncology supercomputer how to prescribe the best treatment plans based on an individual cancer patient’s personal history.

The AI system is not only using massive amounts of data and existing protocols, it’s collecting empirical information from the doctors as well. The result is a holistic treatment model that combines human input with the computer’s massive ability to gather and analyze data.

Personalization in healthcare is nowhere as important as in the treatment of deadly diseases like cancer, where information like results from tests, pathology screenings, images, and even genetic mutations can be unique to each individual.Once IBM Watson for Oncology receives this information about the patient, it can extract the correct treatment plan based on thousands of data points from other cancer treatment results and protocols.

Although IBM Watson is still in a testing phase in most U.S. hospitals, it will soon be an important way for doctors to develop treatment plans more quickly and efficiently, and with a great deal more data behind it.

#4 Trend – Robots and Monitoring

One area of concern as our population ages is the desire of seniors to stay in their homes. Caretakers also want this to happen, but safety is a major concern. One huge trend for AI in healthcare is the use of robotic monitoring tools to send information to doctors, nurses and other care staff.

Rather than having to go to a multitude of doctor appointments, robotic agents and sensors can send data straight to the doctors’ office, helping detect changes in blood pressure, heart rate, or even mood. For care teams, passive sensors in the floor can detect a possible fall, or a lack of movement, within a prescribed period of time.

Although experts believe we’re about two decades from robots that can actually develop empathy and more personal interactions, this application of AI in healthcare is already helping seniors.

CareCoach is a company that deploys robot pets to provide companionship and also monitor seniors as they go about their daily lives. These “pets” can detect important changes in the person’s activity and submit those changes back to the doctor.

In addition to monitoring, the pets can assist in motivating important behaviors like taking medicine as prescribed, and getting enough to eat.

As our population ages, motivating good behaviors will be an important part of health engagement. Artificial intelligence in healthcare in the form of robotic “agents” could play a powerful role in maintaining quality of life for our seniors.

The applications of AI in healthcare are enormous and its implications will be seen in many more instances than the ones we’ve described above. And yet those four areas – screening, diagnosis, treatment, and monitoring – are most likely the earliest to see big benefits from machines that are learning more about us every single day.